Adult drink driving rates [updated]

And here I thought that the slippery slope was the thing to be less worried about. A youth drink driving limit of zero has the same risk as an adult rate of 0.05, and now they're talking about dropping the adult rate to 0.05.

The best possible public safety argument on this is that, if you get to 0.06, you're far more likely to hit 0.12 than if you stop at 0.04. And, of course, that reduces to not allowing any drinking at all because the first drink is the gateway drug to the second.

Look at the graph on my prior post. That curve really starts getting sharp for the over 30s around 0.11; for the 20-30s, around .08-.1. Last time I'd looked at the accident stats, most folks involved in the bad crashes were well over 0.08 and few were in the 0.05-0.08 range. Why? Because folks in the 0.05-0.08 range don't have any substantial impairment in their driving but are paranoid about being pulled over in case they're just over the limit, so they drive more carefully. Where did I see those numbers...ah yes. Farrar pointed to them when this was first suggested last year.

And, of course, nothing shows that the accidents in any part of that range are causally due to alcohol. You'd need the number of accidents relative to the proportion of drivers on the road with the stated levels of alcohol. With the number of folks who'll drive in the <80 mg/100mL range, I'd be very surprised if the 4 deaths are more than we'd expect among sober drivers (sober people do crash too). Where do we see the big increase in deaths? Above 0.12, just as we'd expect from the graph I'd put up earlier.

Mr Joyce said Cabinet had not yet decided about whether or not to reduce the current adult blood alcohol limit from 80mg/100ml (0.08) to 50mg/100ml (0.05).

"This is a very finely balanced argument and we need to ensure that New Zealanders understand the difference between 0.05 and 0.08 and what impact a change would have on the road toll," Mr Joyce said.

Most New Zealanders, when asked, Mr Joyce said, agreed the limit should be lower but were split when asked if it should be lowered to 0.05.

What difference would it have on the road toll? Two drivers with blood alcohol between 0.05 and 0.08 died in accidents in 2007. Two. And we can't even prove that alcohol there was causal.

Update: MacDoctor has a nice series of posts from the last couple of years making the same point. See here here here here here. One nice bit:

I know I will be called reactionary for this, but it seems plain to me that the answer to drunk driving is not to fiddle with blood alcohol limits. The answer is to get draconian with drunk drivers. I have heard of a person being fined and their license suspended for two years for their sixth offense – an offense that included driving without a license. What was the point of that? Why is drunk driving such a joke to us Kiwis? There is no point in the law being a mere punishment, it needs to be a very scary deterrent.

Indeed. Dropping the blood alcohol limit while using a wet bus ticket to slap folks driving at 0.12 for the tenth time is asinine.